diff options
23 files changed, 139 insertions, 127 deletions
diff --git a/apps/plugins/mpegplayer/libmpeg2/idct_arm.S b/apps/plugins/mpegplayer/libmpeg2/idct_arm.S index 97a87a8b59..90eb5031c7 100644 --- a/apps/plugins/mpegplayer/libmpeg2/idct_arm.S +++ b/apps/plugins/mpegplayer/libmpeg2/idct_arm.S | |||
@@ -43,8 +43,8 @@ | |||
43 | ldrsh r7, [r0, #12] /* d2 */ | 43 | ldrsh r7, [r0, #12] /* d2 */ |
44 | ldrsh r8, [r0, #14] /* d3 */ | 44 | ldrsh r8, [r0, #14] /* d3 */ |
45 | orrs r9, r2, r3 | 45 | orrs r9, r2, r3 |
46 | orreqs r9, r4, r5 | 46 | orrseq r9, r4, r5 |
47 | orreqs r9, r6, r7 | 47 | orrseq r9, r6, r7 |
48 | cmpeq r8, #0 | 48 | cmpeq r8, #0 |
49 | bne 2f | 49 | bne 2f |
50 | mov r1, r1, asl #15 | 50 | mov r1, r1, asl #15 |
@@ -320,7 +320,7 @@ mpeg2_idct_copy: | |||
320 | mpeg2_idct_add: | 320 | mpeg2_idct_add: |
321 | cmp r0, #129 | 321 | cmp r0, #129 |
322 | mov r0, r1 | 322 | mov r0, r1 |
323 | ldreqsh r1, [r0, #0] | 323 | ldrsheq r1, [r0, #0] |
324 | bne 1f | 324 | bne 1f |
325 | and r1, r1, #0x70 | 325 | and r1, r1, #0x70 |
326 | cmp r1, #0x40 | 326 | cmp r1, #0x40 |
diff --git a/apps/plugins/mpegplayer/libmpeg2/idct_armv6.S b/apps/plugins/mpegplayer/libmpeg2/idct_armv6.S index dc53cbd7bd..a259721410 100644 --- a/apps/plugins/mpegplayer/libmpeg2/idct_armv6.S +++ b/apps/plugins/mpegplayer/libmpeg2/idct_armv6.S | |||
@@ -19,6 +19,7 @@ | |||
19 | * | 19 | * |
20 | ****************************************************************************/ | 20 | ****************************************************************************/ |
21 | 21 | ||
22 | #include "config.h" | ||
22 | 23 | ||
23 | .global mpeg2_idct_copy | 24 | .global mpeg2_idct_copy |
24 | .type mpeg2_idct_copy, %function | 25 | .type mpeg2_idct_copy, %function |
@@ -228,7 +229,7 @@ mpeg2_idct_copy: | |||
228 | mpeg2_idct_add: | 229 | mpeg2_idct_add: |
229 | cmp r0, #129 | 230 | cmp r0, #129 |
230 | mov r0, r1 | 231 | mov r0, r1 |
231 | ldreqsh r1, [r0, #0] | 232 | ldrsheq r1, [r0, #0] |
232 | bne 1f | 233 | bne 1f |
233 | and r1, r1, #0x70 | 234 | and r1, r1, #0x70 |
234 | cmp r1, #0x40 | 235 | cmp r1, #0x40 |
@@ -260,7 +261,7 @@ mpeg2_idct_add: | |||
260 | strd r4, [r1] @ r4, r5 | 261 | strd r4, [r1] @ r4, r5 |
261 | add r1, r1, r2 | 262 | add r1, r1, r2 |
262 | cmp r0, r3 | 263 | cmp r0, r3 |
263 | ldrlod r8, [r1] @ r8, r9 | 264 | ldrdlo r8, [r1] @ r8, r9 |
264 | blo 2b | 265 | blo 2b |
265 | 266 | ||
266 | ldmfd sp!, {r4-r11, pc} | 267 | ldmfd sp!, {r4-r11, pc} |
@@ -291,7 +292,7 @@ mpeg2_idct_add: | |||
291 | strd r0, [r2] @ r0, r1 | 292 | strd r0, [r2] @ r0, r1 |
292 | add r2, r2, r3 | 293 | add r2, r2, r3 |
293 | cmp r2, r12 | 294 | cmp r2, r12 |
294 | ldrlod r0, [r2] @ r0, r1 | 295 | ldrdlo r0, [r2] @ r0, r1 |
295 | blo 4b | 296 | blo 4b |
296 | 297 | ||
297 | ldmfd sp!, {r4, pc} | 298 | ldmfd sp!, {r4, pc} |
diff --git a/apps/plugins/mpegplayer/mpeg_misc.h b/apps/plugins/mpegplayer/mpeg_misc.h index e04db0e19d..68ee8cac3c 100644 --- a/apps/plugins/mpegplayer/mpeg_misc.h +++ b/apps/plugins/mpegplayer/mpeg_misc.h | |||
@@ -53,12 +53,13 @@ enum state_enum | |||
53 | #define CMP_3_CONST(_a, _b) \ | 53 | #define CMP_3_CONST(_a, _b) \ |
54 | ({ int _x; \ | 54 | ({ int _x; \ |
55 | asm volatile ( \ | 55 | asm volatile ( \ |
56 | ".syntax unified \n" \ | ||
56 | "ldrb %[x], [%[a], #0] \n" \ | 57 | "ldrb %[x], [%[a], #0] \n" \ |
57 | "eors %[x], %[x], %[b0] \n" \ | 58 | "eors %[x], %[x], %[b0] \n" \ |
58 | "ldreqb %[x], [%[a], #1] \n" \ | 59 | "ldrbeq %[x], [%[a], #1] \n" \ |
59 | "eoreqs %[x], %[x], %[b1] \n" \ | 60 | "eorseq %[x], %[x], %[b1] \n" \ |
60 | "ldreqb %[x], [%[a], #2] \n" \ | 61 | "ldrbeq %[x], [%[a], #2] \n" \ |
61 | "eoreqs %[x], %[x], %[b2] \n" \ | 62 | "eorseq %[x], %[x], %[b2] \n" \ |
62 | : [x]"=&r"(_x) \ | 63 | : [x]"=&r"(_x) \ |
63 | : [a]"r"(_a), \ | 64 | : [a]"r"(_a), \ |
64 | [b0]"i"(((_b) >> 24) & 0xff), \ | 65 | [b0]"i"(((_b) >> 24) & 0xff), \ |
@@ -70,14 +71,15 @@ enum state_enum | |||
70 | #define CMP_4_CONST(_a, _b) \ | 71 | #define CMP_4_CONST(_a, _b) \ |
71 | ({ int _x; \ | 72 | ({ int _x; \ |
72 | asm volatile ( \ | 73 | asm volatile ( \ |
74 | ".syntax unified \n" \ | ||
73 | "ldrb %[x], [%[a], #0] \n" \ | 75 | "ldrb %[x], [%[a], #0] \n" \ |
74 | "eors %[x], %[x], %[b0] \n" \ | 76 | "eors %[x], %[x], %[b0] \n" \ |
75 | "ldreqb %[x], [%[a], #1] \n" \ | 77 | "ldrbeq %[x], [%[a], #1] \n" \ |
76 | "eoreqs %[x], %[x], %[b1] \n" \ | 78 | "eorseq %[x], %[x], %[b1] \n" \ |
77 | "ldreqb %[x], [%[a], #2] \n" \ | 79 | "ldrbeq %[x], [%[a], #2] \n" \ |
78 | "eoreqs %[x], %[x], %[b2] \n" \ | 80 | "eorseq %[x], %[x], %[b2] \n" \ |
79 | "ldreqb %[x], [%[a], #3] \n" \ | 81 | "ldrbeq %[x], [%[a], #3] \n" \ |
80 | "eoreqs %[x], %[x], %[b3] \n" \ | 82 | "eorseq %[x], %[x], %[b3] \n" \ |
81 | : [x]"=&r"(_x) \ | 83 | : [x]"=&r"(_x) \ |
82 | : [a]"r"(_a), \ | 84 | : [a]"r"(_a), \ |
83 | [b0]"i"(((_b) >> 24) & 0xff), \ | 85 | [b0]"i"(((_b) >> 24) & 0xff), \ |
diff --git a/apps/recorder/jpeg_idct_arm.S b/apps/recorder/jpeg_idct_arm.S index e7eb4b87f1..1f2603da1b 100644 --- a/apps/recorder/jpeg_idct_arm.S +++ b/apps/recorder/jpeg_idct_arm.S | |||
@@ -410,7 +410,7 @@ jpeg_idct8v: | |||
410 | #if ARM_ARCH < 5 | 410 | #if ARM_ARCH < 5 |
411 | mov r8, r4, lsl #16 | 411 | mov r8, r4, lsl #16 |
412 | orrs r9, r6, r7 | 412 | orrs r9, r6, r7 |
413 | orreqs r9, r5, r4, lsr #16 | 413 | orrseq r9, r5, r4, lsr #16 |
414 | bne 2f | 414 | bne 2f |
415 | mov r8, r8, asr #14 | 415 | mov r8, r8, asr #14 |
416 | strh r8, [r2] | 416 | strh r8, [r2] |
@@ -505,7 +505,7 @@ jpeg_idct8v: | |||
505 | #else /* ARMv5+ */ | 505 | #else /* ARMv5+ */ |
506 | mov r12, r4, lsl #16 | 506 | mov r12, r4, lsl #16 |
507 | orrs r9, r6, r7 | 507 | orrs r9, r6, r7 |
508 | orreqs r9, r5, r4, lsr #16 | 508 | orrseq r9, r5, r4, lsr #16 |
509 | bne 2f | 509 | bne 2f |
510 | mov r12, r12, asr #14 | 510 | mov r12, r12, asr #14 |
511 | strh r12, [r2] | 511 | strh r12, [r2] |
@@ -615,7 +615,7 @@ jpeg_idct8h: | |||
615 | #if ARM_ARCH < 5 | 615 | #if ARM_ARCH < 5 |
616 | add r8, r14, r4, lsl #16 | 616 | add r8, r14, r4, lsl #16 |
617 | orrs r9, r6, r7 | 617 | orrs r9, r6, r7 |
618 | orreqs r9, r5, r4, lsr #16 | 618 | orrseq r9, r5, r4, lsr #16 |
619 | bne 2f | 619 | bne 2f |
620 | mov r8, r8, asr #21 | 620 | mov r8, r8, asr #21 |
621 | cmp r8, #255 | 621 | cmp r8, #255 |
@@ -727,7 +727,7 @@ jpeg_idct8h: | |||
727 | #else /* ARMv5+ */ | 727 | #else /* ARMv5+ */ |
728 | add r12, r14, r4, lsl #16 | 728 | add r12, r14, r4, lsl #16 |
729 | orrs r9, r6, r7 | 729 | orrs r9, r6, r7 |
730 | orreqs r9, r5, r4, lsr #16 | 730 | orrseq r9, r5, r4, lsr #16 |
731 | bne 2f | 731 | bne 2f |
732 | mov r12, r12, asr #21 | 732 | mov r12, r12, asr #21 |
733 | cmp r12, #255 | 733 | cmp r12, #255 |
@@ -835,7 +835,7 @@ jpeg_idct8v: | |||
835 | 1: | 835 | 1: |
836 | ldmia r0!, { r4-r7 } | 836 | ldmia r0!, { r4-r7 } |
837 | orrs r9, r6, r7 | 837 | orrs r9, r6, r7 |
838 | orreqs r9, r5, r4, lsr #16 | 838 | orrseq r9, r5, r4, lsr #16 |
839 | bne 2f | 839 | bne 2f |
840 | mov r4, r4, lsl #2 | 840 | mov r4, r4, lsl #2 |
841 | strh r4, [r2] | 841 | strh r4, [r2] |
@@ -939,7 +939,7 @@ jpeg_idct8h: | |||
939 | ldmia r0!, { r4-r7 } | 939 | ldmia r0!, { r4-r7 } |
940 | sadd16 r4, r4, r14 | 940 | sadd16 r4, r4, r14 |
941 | orrs r9, r6, r7 | 941 | orrs r9, r6, r7 |
942 | orreqs r9, r5, r4, lsr #16 | 942 | orrseq r9, r5, r4, lsr #16 |
943 | bne 2f | 943 | bne 2f |
944 | sxth r4, r4 | 944 | sxth r4, r4 |
945 | usat r4, #8, r4, asr #5 | 945 | usat r4, #8, r4, asr #5 |
diff --git a/firmware/asm/arm/corelock.c b/firmware/asm/arm/corelock.c index b36a40b45b..07ec77a60e 100644 --- a/firmware/asm/arm/corelock.c +++ b/firmware/asm/arm/corelock.c | |||
@@ -61,6 +61,7 @@ int corelock_try_lock(struct corelock *cl) | |||
61 | 61 | ||
62 | /* Relies on the fact that core IDs are complementary bitmasks (0x55,0xaa) */ | 62 | /* Relies on the fact that core IDs are complementary bitmasks (0x55,0xaa) */ |
63 | asm volatile ( | 63 | asm volatile ( |
64 | ".syntax unified \n" | ||
64 | "mov r1, %[id] \n" /* r1 = PROCESSOR_ID */ | 65 | "mov r1, %[id] \n" /* r1 = PROCESSOR_ID */ |
65 | "ldrb r1, [r1] \n" | 66 | "ldrb r1, [r1] \n" |
66 | "strb r1, [%[cl], r1, lsr #7] \n" /* cl->myl[core] = core */ | 67 | "strb r1, [%[cl], r1, lsr #7] \n" /* cl->myl[core] = core */ |
@@ -71,7 +72,7 @@ int corelock_try_lock(struct corelock *cl) | |||
71 | "bne 1f \n" /* yes? lock acquired */ | 72 | "bne 1f \n" /* yes? lock acquired */ |
72 | "ldrb %[rv], [%[cl], #2] \n" /* || cl->turn == core? */ | 73 | "ldrb %[rv], [%[cl], #2] \n" /* || cl->turn == core? */ |
73 | "ands %[rv], %[rv], r1 \n" | 74 | "ands %[rv], %[rv], r1 \n" |
74 | "streqb %[rv], [%[cl], r1, lsr #7] \n" /* if not, cl->myl[core] = 0 */ | 75 | "strbeq %[rv], [%[cl], r1, lsr #7] \n" /* if not, cl->myl[core] = 0 */ |
75 | "1: \n" /* Done */ | 76 | "1: \n" /* Done */ |
76 | : [rv] "=r"(rval) | 77 | : [rv] "=r"(rval) |
77 | : [id] "i" (&PROCESSOR_ID), [cl] "r" (cl) | 78 | : [id] "i" (&PROCESSOR_ID), [cl] "r" (cl) |
diff --git a/firmware/asm/arm/lcd-as-memframe.S b/firmware/asm/arm/lcd-as-memframe.S index 52ab0447c2..d42b2a920d 100644 --- a/firmware/asm/arm/lcd-as-memframe.S +++ b/firmware/asm/arm/lcd-as-memframe.S | |||
@@ -91,9 +91,9 @@ lcd_copy_buffer_rect: @ | |||
91 | stmia r0!, { r6-r12, r14 } @ | 91 | stmia r0!, { r6-r12, r14 } @ |
92 | bgt 30b @ octword loop @ | 92 | bgt 30b @ octword loop @ |
93 | 40: @ finish line @ | 93 | 40: @ finish line @ |
94 | ldreqh r6, [r1], #2 @ finish last halfword if eq ... | 94 | ldrheq r6, [r1], #2 @ finish last halfword if eq ... |
95 | add r1, r1, r4, lsl #1 @ | 95 | add r1, r1, r4, lsl #1 @ |
96 | streqh r6, [r0], #2 @ ... | 96 | strheq r6, [r0], #2 @ ... |
97 | add r0, r0, r4, lsl #1 @ | 97 | add r0, r0, r4, lsl #1 @ |
98 | subs r3, r3, #1 @ next line | 98 | subs r3, r3, #1 @ next line |
99 | bgt 10b @ copy line @ | 99 | bgt 10b @ copy line @ |
diff --git a/firmware/asm/arm/memcpy.S b/firmware/asm/arm/memcpy.S index 83d43293e6..86fc6b7930 100644 --- a/firmware/asm/arm/memcpy.S +++ b/firmware/asm/arm/memcpy.S | |||
@@ -99,22 +99,22 @@ memcpy: | |||
99 | 7: ldmfd sp!, {r5 - r8} | 99 | 7: ldmfd sp!, {r5 - r8} |
100 | 100 | ||
101 | 8: movs r2, r2, lsl #31 | 101 | 8: movs r2, r2, lsl #31 |
102 | ldrneb r3, [r1], #1 | 102 | ldrbne r3, [r1], #1 |
103 | ldrcsb r4, [r1], #1 | 103 | ldrbcs r4, [r1], #1 |
104 | ldrcsb ip, [r1] | 104 | ldrbcs ip, [r1] |
105 | strneb r3, [r0], #1 | 105 | strbne r3, [r0], #1 |
106 | strcsb r4, [r0], #1 | 106 | strbcs r4, [r0], #1 |
107 | strcsb ip, [r0] | 107 | strbcs ip, [r0] |
108 | 108 | ||
109 | ldmpc regs="r0, r4" | 109 | ldmpc regs="r0, r4" |
110 | 110 | ||
111 | 9: rsb ip, ip, #4 | 111 | 9: rsb ip, ip, #4 |
112 | cmp ip, #2 | 112 | cmp ip, #2 |
113 | ldrgtb r3, [r1], #1 | 113 | ldrbgt r3, [r1], #1 |
114 | ldrgeb r4, [r1], #1 | 114 | ldrbge r4, [r1], #1 |
115 | ldrb lr, [r1], #1 | 115 | ldrb lr, [r1], #1 |
116 | strgtb r3, [r0], #1 | 116 | strbgt r3, [r0], #1 |
117 | strgeb r4, [r0], #1 | 117 | strbge r4, [r0], #1 |
118 | subs r2, r2, ip | 118 | subs r2, r2, ip |
119 | strb lr, [r0], #1 | 119 | strb lr, [r0], #1 |
120 | blt 8b | 120 | blt 8b |
diff --git a/firmware/asm/arm/memmove.S b/firmware/asm/arm/memmove.S index d8cab048be..e5c9b42928 100644 --- a/firmware/asm/arm/memmove.S +++ b/firmware/asm/arm/memmove.S | |||
@@ -106,20 +106,20 @@ memmove: | |||
106 | 7: ldmfd sp!, {r5 - r8} | 106 | 7: ldmfd sp!, {r5 - r8} |
107 | 107 | ||
108 | 8: movs r2, r2, lsl #31 | 108 | 8: movs r2, r2, lsl #31 |
109 | ldrneb r3, [r1, #-1]! | 109 | ldrbne r3, [r1, #-1]! |
110 | ldrcsb r4, [r1, #-1]! | 110 | ldrbcs r4, [r1, #-1]! |
111 | ldrcsb ip, [r1, #-1] | 111 | ldrbcs ip, [r1, #-1] |
112 | strneb r3, [r0, #-1]! | 112 | strbne r3, [r0, #-1]! |
113 | strcsb r4, [r0, #-1]! | 113 | strbcs r4, [r0, #-1]! |
114 | strcsb ip, [r0, #-1] | 114 | strbcs ip, [r0, #-1] |
115 | ldmpc regs="r0, r4" | 115 | ldmpc regs="r0, r4" |
116 | 116 | ||
117 | 9: cmp ip, #2 | 117 | 9: cmp ip, #2 |
118 | ldrgtb r3, [r1, #-1]! | 118 | ldrbgt r3, [r1, #-1]! |
119 | ldrgeb r4, [r1, #-1]! | 119 | ldrbge r4, [r1, #-1]! |
120 | ldrb lr, [r1, #-1]! | 120 | ldrb lr, [r1, #-1]! |
121 | strgtb r3, [r0, #-1]! | 121 | strbgt r3, [r0, #-1]! |
122 | strgeb r4, [r0, #-1]! | 122 | strbge r4, [r0, #-1]! |
123 | subs r2, r2, ip | 123 | subs r2, r2, ip |
124 | strb lr, [r0, #-1]! | 124 | strb lr, [r0, #-1]! |
125 | blt 8b | 125 | blt 8b |
diff --git a/firmware/asm/arm/memset.S b/firmware/asm/arm/memset.S index 64cd95cc9e..d727f2a5ec 100644 --- a/firmware/asm/arm/memset.S +++ b/firmware/asm/arm/memset.S | |||
@@ -34,8 +34,8 @@ | |||
34 | 1: cmp r2, #4 @ 1 do we have enough | 34 | 1: cmp r2, #4 @ 1 do we have enough |
35 | blt 5f @ 1 bytes to align with? | 35 | blt 5f @ 1 bytes to align with? |
36 | cmp r3, #2 @ 1 | 36 | cmp r3, #2 @ 1 |
37 | strgtb r1, [r0, #-1]! @ 1 | 37 | strbgt r1, [r0, #-1]! @ 1 |
38 | strgeb r1, [r0, #-1]! @ 1 | 38 | strbge r1, [r0, #-1]! @ 1 |
39 | strb r1, [r0, #-1]! @ 1 | 39 | strb r1, [r0, #-1]! @ 1 |
40 | sub r2, r2, r3 @ 1 r2 = r2 - r3 | 40 | sub r2, r2, r3 @ 1 r2 = r2 - r3 |
41 | b 2f | 41 | b 2f |
@@ -65,24 +65,24 @@ memset: | |||
65 | mov lr, r1 | 65 | mov lr, r1 |
66 | 66 | ||
67 | 3: subs r2, r2, #64 | 67 | 3: subs r2, r2, #64 |
68 | stmgedb r0!, {r1, r3, ip, lr} @ 64 bytes at a time. | 68 | stmdbge r0!, {r1, r3, ip, lr} @ 64 bytes at a time. |
69 | stmgedb r0!, {r1, r3, ip, lr} | 69 | stmdbge r0!, {r1, r3, ip, lr} |
70 | stmgedb r0!, {r1, r3, ip, lr} | 70 | stmdbge r0!, {r1, r3, ip, lr} |
71 | stmgedb r0!, {r1, r3, ip, lr} | 71 | stmdbge r0!, {r1, r3, ip, lr} |
72 | bgt 3b | 72 | bgt 3b |
73 | ldrpc cond=eq @ Now <64 bytes to go. | 73 | ldrpc cond=eq @ Now <64 bytes to go. |
74 | /* | 74 | /* |
75 | * No need to correct the count; we're only testing bits from now on | 75 | * No need to correct the count; we're only testing bits from now on |
76 | */ | 76 | */ |
77 | tst r2, #32 | 77 | tst r2, #32 |
78 | stmnedb r0!, {r1, r3, ip, lr} | 78 | stmdbne r0!, {r1, r3, ip, lr} |
79 | stmnedb r0!, {r1, r3, ip, lr} | 79 | stmdbne r0!, {r1, r3, ip, lr} |
80 | tst r2, #16 | 80 | tst r2, #16 |
81 | stmnedb r0!, {r1, r3, ip, lr} | 81 | stmdbne r0!, {r1, r3, ip, lr} |
82 | ldr lr, [sp], #4 | 82 | ldr lr, [sp], #4 |
83 | 83 | ||
84 | 5: tst r2, #8 | 84 | 5: tst r2, #8 |
85 | stmnedb r0!, {r1, r3} | 85 | stmdbne r0!, {r1, r3} |
86 | tst r2, #4 | 86 | tst r2, #4 |
87 | strne r1, [r0, #-4]! | 87 | strne r1, [r0, #-4]! |
88 | /* | 88 | /* |
@@ -90,10 +90,10 @@ memset: | |||
90 | * may have an unaligned pointer as well. | 90 | * may have an unaligned pointer as well. |
91 | */ | 91 | */ |
92 | 6: tst r2, #2 | 92 | 6: tst r2, #2 |
93 | strneb r1, [r0, #-1]! | 93 | strbne r1, [r0, #-1]! |
94 | strneb r1, [r0, #-1]! | 94 | strbne r1, [r0, #-1]! |
95 | tst r2, #1 | 95 | tst r2, #1 |
96 | strneb r1, [r0, #-1]! | 96 | strbne r1, [r0, #-1]! |
97 | bx lr | 97 | bx lr |
98 | .end: | 98 | .end: |
99 | .size memset,.end-memset | 99 | .size memset,.end-memset |
diff --git a/firmware/asm/arm/memset16.S b/firmware/asm/arm/memset16.S index 5c787b1bed..226eac39e1 100644 --- a/firmware/asm/arm/memset16.S +++ b/firmware/asm/arm/memset16.S | |||
@@ -35,7 +35,7 @@ | |||
35 | memset16: | 35 | memset16: |
36 | tst r0, #2 @ unaligned? | 36 | tst r0, #2 @ unaligned? |
37 | cmpne r2, #0 | 37 | cmpne r2, #0 |
38 | strneh r1, [r0], #2 @ store one halfword to align | 38 | strhne r1, [r0], #2 @ store one halfword to align |
39 | subne r2, r2, #1 | 39 | subne r2, r2, #1 |
40 | 40 | ||
41 | /* | 41 | /* |
@@ -54,29 +54,29 @@ memset16: | |||
54 | mov lr, r1 | 54 | mov lr, r1 |
55 | 55 | ||
56 | 2: subs r2, r2, #32 | 56 | 2: subs r2, r2, #32 |
57 | stmgeia r0!, {r1, r3, ip, lr} @ 64 bytes at a time. | 57 | stmiage r0!, {r1, r3, ip, lr} @ 64 bytes at a time. |
58 | stmgeia r0!, {r1, r3, ip, lr} | 58 | stmiage r0!, {r1, r3, ip, lr} |
59 | stmgeia r0!, {r1, r3, ip, lr} | 59 | stmiage r0!, {r1, r3, ip, lr} |
60 | stmgeia r0!, {r1, r3, ip, lr} | 60 | stmiage r0!, {r1, r3, ip, lr} |
61 | bgt 2b | 61 | bgt 2b |
62 | ldrpc cond=eq @ Now <64 bytes to go. | 62 | ldrpc cond=eq @ Now <64 bytes to go. |
63 | /* | 63 | /* |
64 | * No need to correct the count; we're only testing bits from now on | 64 | * No need to correct the count; we're only testing bits from now on |
65 | */ | 65 | */ |
66 | tst r2, #16 | 66 | tst r2, #16 |
67 | stmneia r0!, {r1, r3, ip, lr} | 67 | stmiane r0!, {r1, r3, ip, lr} |
68 | stmneia r0!, {r1, r3, ip, lr} | 68 | stmiane r0!, {r1, r3, ip, lr} |
69 | tst r2, #8 | 69 | tst r2, #8 |
70 | stmneia r0!, {r1, r3, ip, lr} | 70 | stmiane r0!, {r1, r3, ip, lr} |
71 | ldr lr, [sp], #4 | 71 | ldr lr, [sp], #4 |
72 | 72 | ||
73 | 4: tst r2, #4 | 73 | 4: tst r2, #4 |
74 | stmneia r0!, {r1, r3} | 74 | stmiane r0!, {r1, r3} |
75 | tst r2, #2 | 75 | tst r2, #2 |
76 | strne r1, [r0], #4 | 76 | strne r1, [r0], #4 |
77 | 77 | ||
78 | tst r2, #1 | 78 | tst r2, #1 |
79 | strneh r1, [r0], #2 | 79 | strhne r1, [r0], #2 |
80 | bx lr | 80 | bx lr |
81 | .end: | 81 | .end: |
82 | .size memset16,.end-memset16 | 82 | .size memset16,.end-memset16 |
diff --git a/firmware/asm/arm/thread.c b/firmware/asm/arm/thread.c index cf685526e3..bd9f950616 100644 --- a/firmware/asm/arm/thread.c +++ b/firmware/asm/arm/thread.c | |||
@@ -73,15 +73,16 @@ static inline void store_context(void* addr) | |||
73 | static inline void load_context(const void* addr) | 73 | static inline void load_context(const void* addr) |
74 | { | 74 | { |
75 | asm volatile( | 75 | asm volatile( |
76 | ".syntax unified \n" | ||
76 | "ldr r0, [%0, #40] \n" /* Load start pointer */ | 77 | "ldr r0, [%0, #40] \n" /* Load start pointer */ |
77 | "cmp r0, #0 \n" /* Check for NULL */ | 78 | "cmp r0, #0 \n" /* Check for NULL */ |
78 | 79 | ||
79 | /* If not already running, jump to start */ | 80 | /* If not already running, jump to start */ |
80 | #if ARM_ARCH == 4 && defined(USE_THUMB) | 81 | #if ARM_ARCH == 4 && defined(USE_THUMB) |
81 | "ldmneia %0, { r0, r12 } \n" | 82 | "ldmiane %0, { r0, r12 } \n" |
82 | "bxne r12 \n" | 83 | "bxne r12 \n" |
83 | #else | 84 | #else |
84 | "ldmneia %0, { r0, pc } \n" | 85 | "ldmiane %0, { r0, pc } \n" |
85 | #endif | 86 | #endif |
86 | 87 | ||
87 | "ldmia %0, { r4-r11, sp, lr } \n" /* Load regs r4 to r14 from context */ | 88 | "ldmia %0, { r4-r11, sp, lr } \n" /* Load regs r4 to r14 from context */ |
diff --git a/firmware/export/config.h b/firmware/export/config.h index 0882cad61c..19ee03b4c7 100644 --- a/firmware/export/config.h +++ b/firmware/export/config.h | |||
@@ -1006,13 +1006,14 @@ Lyre prototype 1 */ | |||
1006 | #endif | 1006 | #endif |
1007 | 1007 | ||
1008 | #if defined(CPU_ARM) && defined(__ASSEMBLER__) | 1008 | #if defined(CPU_ARM) && defined(__ASSEMBLER__) |
1009 | .syntax unified | ||
1009 | /* ARMv4T doesn't switch the T bit when popping pc directly, we must use BX */ | 1010 | /* ARMv4T doesn't switch the T bit when popping pc directly, we must use BX */ |
1010 | .macro ldmpc cond="", order="ia", regs | 1011 | .macro ldmpc cond="", order="ia", regs |
1011 | #if ARM_ARCH == 4 && defined(USE_THUMB) | 1012 | #if ARM_ARCH == 4 && defined(USE_THUMB) |
1012 | ldm\cond\order sp!, { \regs, lr } | 1013 | ldm\order\cond sp!, { \regs, lr } |
1013 | bx\cond lr | 1014 | bx\cond lr |
1014 | #else | 1015 | #else |
1015 | ldm\cond\order sp!, { \regs, pc } | 1016 | ldm\order\cond sp!, { \regs, pc } |
1016 | #endif | 1017 | #endif |
1017 | .endm | 1018 | .endm |
1018 | .macro ldrpc cond="" | 1019 | .macro ldrpc cond="" |
diff --git a/firmware/target/arm/ata-as-arm.S b/firmware/target/arm/ata-as-arm.S index 101bc4dcc1..16c2928bf1 100644 --- a/firmware/target/arm/ata-as-arm.S +++ b/firmware/target/arm/ata-as-arm.S | |||
@@ -139,9 +139,9 @@ copy_read_sectors: | |||
139 | .r_end2_u: | 139 | .r_end2_u: |
140 | 140 | ||
141 | tst r1, #1 /* one halfword left? */ | 141 | tst r1, #1 /* one halfword left? */ |
142 | ldrneh r4, [r2] | 142 | ldrhne r4, [r2] |
143 | orrne r3, r3, r4, lsl #8 | 143 | orrne r3, r3, r4, lsl #8 |
144 | strneh r3, [r0], #2 | 144 | strhne r3, [r0], #2 |
145 | movne r3, r4, lsr #8 | 145 | movne r3, r4, lsr #8 |
146 | 146 | ||
147 | strb r3, [r0], #1 /* store final byte */ | 147 | strb r3, [r0], #1 /* store final byte */ |
@@ -151,8 +151,8 @@ copy_read_sectors: | |||
151 | /* 16-bit aligned */ | 151 | /* 16-bit aligned */ |
152 | .r_aligned: | 152 | .r_aligned: |
153 | tst r0, #2 /* 32 bit aligned? */ | 153 | tst r0, #2 /* 32 bit aligned? */ |
154 | ldrneh r3, [r2] /* no: read first halfword */ | 154 | ldrhne r3, [r2] /* no: read first halfword */ |
155 | strneh r3, [r0], #2 /* store */ | 155 | strhne r3, [r0], #2 /* store */ |
156 | subne r1, r1, #1 /* one halfword taken */ | 156 | subne r1, r1, #1 /* one halfword taken */ |
157 | 157 | ||
158 | sub r1, r1, #8 /* adjust for zero-check and doing 8 halfwords/loop */ | 158 | sub r1, r1, #8 /* adjust for zero-check and doing 8 halfwords/loop */ |
@@ -186,14 +186,14 @@ copy_read_sectors: | |||
186 | .r_end4_a: | 186 | .r_end4_a: |
187 | 187 | ||
188 | tst r1, #2 /* 2 or more halfwords left? */ | 188 | tst r1, #2 /* 2 or more halfwords left? */ |
189 | ldrneh r3, [r2] | 189 | ldrhne r3, [r2] |
190 | ldrneh r4, [r2] | 190 | ldrhne r4, [r2] |
191 | orrne r3, r3, r4, lsl #16 | 191 | orrne r3, r3, r4, lsl #16 |
192 | strne r3, [r0], #4 | 192 | strne r3, [r0], #4 |
193 | 193 | ||
194 | tst r1, #1 /* one halfword left? */ | 194 | tst r1, #1 /* one halfword left? */ |
195 | ldrneh r3, [r2] | 195 | ldrhne r3, [r2] |
196 | strneh r3, [r0], #2 | 196 | strhne r3, [r0], #2 |
197 | 197 | ||
198 | ldmpc regs=r4-r5 | 198 | ldmpc regs=r4-r5 |
199 | 199 | ||
@@ -291,9 +291,9 @@ copy_write_sectors: | |||
291 | .w_end2_u: | 291 | .w_end2_u: |
292 | 292 | ||
293 | tst r1, #1 /* one halfword left? */ | 293 | tst r1, #1 /* one halfword left? */ |
294 | ldrneh r4, [r0], #2 | 294 | ldrhne r4, [r0], #2 |
295 | orrne r3, r3, r4, lsl #8 | 295 | orrne r3, r3, r4, lsl #8 |
296 | strneh r3, [r2] | 296 | strhne r3, [r2] |
297 | movne r3, r3, lsr #16 | 297 | movne r3, r3, lsr #16 |
298 | 298 | ||
299 | ldrb r4, [r0], #1 /* load final byte */ | 299 | ldrb r4, [r0], #1 /* load final byte */ |
@@ -305,8 +305,8 @@ copy_write_sectors: | |||
305 | /* 16-bit aligned */ | 305 | /* 16-bit aligned */ |
306 | .w_aligned: | 306 | .w_aligned: |
307 | tst r0, #2 /* 32 bit aligned? */ | 307 | tst r0, #2 /* 32 bit aligned? */ |
308 | ldrneh r3, [r0], #2 /* no: load first halfword */ | 308 | ldrhne r3, [r0], #2 /* no: load first halfword */ |
309 | strneh r3, [r2] /* write */ | 309 | strhne r3, [r2] /* write */ |
310 | subne r1, r1, #1 /* one halfword taken */ | 310 | subne r1, r1, #1 /* one halfword taken */ |
311 | 311 | ||
312 | sub r1, r1, #8 /* adjust for zero-check and doing 8 halfwords/loop */ | 312 | sub r1, r1, #8 /* adjust for zero-check and doing 8 halfwords/loop */ |
@@ -341,13 +341,13 @@ copy_write_sectors: | |||
341 | 341 | ||
342 | tst r1, #2 /* 2 or more halfwords left? */ | 342 | tst r1, #2 /* 2 or more halfwords left? */ |
343 | ldrne r3, [r0], #4 | 343 | ldrne r3, [r0], #4 |
344 | strneh r3, [r2] | 344 | strhne r3, [r2] |
345 | movne r3, r3, lsr #16 | 345 | movne r3, r3, lsr #16 |
346 | strneh r3, [r2] | 346 | strhne r3, [r2] |
347 | 347 | ||
348 | tst r1, #1 /* one halfword left? */ | 348 | tst r1, #1 /* one halfword left? */ |
349 | ldrneh r3, [r0], #2 | 349 | ldrhne r3, [r0], #2 |
350 | strneh r3, [r2] | 350 | strhne r3, [r2] |
351 | 351 | ||
352 | ldmpc regs=r4-r5 | 352 | ldmpc regs=r4-r5 |
353 | 353 | ||
diff --git a/firmware/target/arm/ipod/video/lcd-as-video.S b/firmware/target/arm/ipod/video/lcd-as-video.S index 47155b8c75..7d6caef448 100644 --- a/firmware/target/arm/ipod/video/lcd-as-video.S +++ b/firmware/target/arm/ipod/video/lcd-as-video.S | |||
@@ -40,24 +40,24 @@ lcd_write_data: /* r1 = pixel count, must be even */ | |||
40 | 40 | ||
41 | subs r1, r1, #16 | 41 | subs r1, r1, #16 |
42 | .loop16: | 42 | .loop16: |
43 | ldmgeia r0!, {r2-r3} | 43 | ldmiage r0!, {r2-r3} |
44 | stmgeia lr, {r2-r3} | 44 | stmiage lr, {r2-r3} |
45 | ldmgeia r0!, {r2-r3} | 45 | ldmiage r0!, {r2-r3} |
46 | stmgeia lr, {r2-r3} | 46 | stmiage lr, {r2-r3} |
47 | ldmgeia r0!, {r2-r3} | 47 | ldmiage r0!, {r2-r3} |
48 | stmgeia lr, {r2-r3} | 48 | stmiage lr, {r2-r3} |
49 | ldmgeia r0!, {r2-r3} | 49 | ldmiage r0!, {r2-r3} |
50 | stmgeia lr, {r2-r3} | 50 | stmiage lr, {r2-r3} |
51 | subges r1, r1, #16 | 51 | subsge r1, r1, #16 |
52 | bge .loop16 | 52 | bge .loop16 |
53 | 53 | ||
54 | /* no need to correct the count, we're just checking bits from now */ | 54 | /* no need to correct the count, we're just checking bits from now */ |
55 | tst r1, #8 | 55 | tst r1, #8 |
56 | ldmneia r0!, {r2-r4, r12} | 56 | ldmiane r0!, {r2-r4, r12} |
57 | stmneia lr, {r2-r4, r12} | 57 | stmiane lr, {r2-r4, r12} |
58 | tst r1, #4 | 58 | tst r1, #4 |
59 | ldmneia r0!, {r2-r3} | 59 | ldmiane r0!, {r2-r3} |
60 | stmneia lr, {r2-r3} | 60 | stmiane lr, {r2-r3} |
61 | tst r1, #2 | 61 | tst r1, #2 |
62 | ldrne r3, [r0], #4 | 62 | ldrne r3, [r0], #4 |
63 | strne r3, [lr] | 63 | strne r3, [lr] |
diff --git a/firmware/target/arm/pcm-telechips.c b/firmware/target/arm/pcm-telechips.c index 336b5626ca..747765d8fb 100644 --- a/firmware/target/arm/pcm-telechips.c +++ b/firmware/target/arm/pcm-telechips.c | |||
@@ -218,6 +218,7 @@ void fiq_handler(void) | |||
218 | * r0-r3 and r12 is a working register. | 218 | * r0-r3 and r12 is a working register. |
219 | */ | 219 | */ |
220 | asm volatile ( | 220 | asm volatile ( |
221 | ".syntax unified \n" | ||
221 | "sub lr, lr, #4 \n" | 222 | "sub lr, lr, #4 \n" |
222 | "stmfd sp!, { r0-r3, lr } \n" /* stack scratch regs and lr */ | 223 | "stmfd sp!, { r0-r3, lr } \n" /* stack scratch regs and lr */ |
223 | "mov r14, #0 \n" /* Was the callback called? */ | 224 | "mov r14, #0 \n" /* Was the callback called? */ |
@@ -251,7 +252,7 @@ void fiq_handler(void) | |||
251 | "stmia r11, { r8-r9 } \n" /* save p and size */ | 252 | "stmia r11, { r8-r9 } \n" /* save p and size */ |
252 | 253 | ||
253 | "cmp r14, #0 \n" /* Callback called? */ | 254 | "cmp r14, #0 \n" /* Callback called? */ |
254 | "ldmeqfd sp!, { r0-r3, pc }^ \n" /* no? -> exit */ | 255 | "ldmfdeq sp!, { r0-r3, pc }^ \n" /* no? -> exit */ |
255 | 256 | ||
256 | "ldr r1, =pcm_play_status_callback \n" | 257 | "ldr r1, =pcm_play_status_callback \n" |
257 | "ldr r1, [r1] \n" | 258 | "ldr r1, [r1] \n" |
@@ -268,7 +269,7 @@ void fiq_handler(void) | |||
268 | "mov lr, pc \n" | 269 | "mov lr, pc \n" |
269 | "ldr pc, =pcm_play_dma_complete_callback \n" | 270 | "ldr pc, =pcm_play_dma_complete_callback \n" |
270 | "cmp r0, #0 \n" /* any more to play? */ | 271 | "cmp r0, #0 \n" /* any more to play? */ |
271 | "ldmneia r11, { r8-r9 } \n" /* load new p and size */ | 272 | "ldmiane r11, { r8-r9 } \n" /* load new p and size */ |
272 | "cmpne r9, #0x0f \n" /* did we actually get enough data? */ | 273 | "cmpne r9, #0x0f \n" /* did we actually get enough data? */ |
273 | "bhi .fill_fifo \n" /* not stop and enough? refill */ | 274 | "bhi .fill_fifo \n" /* not stop and enough? refill */ |
274 | "ldmfd sp!, { r0-r3, pc }^ \n" /* exit */ | 275 | "ldmfd sp!, { r0-r3, pc }^ \n" /* exit */ |
diff --git a/firmware/target/arm/pp/pcm-pp.c b/firmware/target/arm/pp/pcm-pp.c index 0d61eb44ff..fd798f0506 100644 --- a/firmware/target/arm/pp/pcm-pp.c +++ b/firmware/target/arm/pp/pcm-pp.c | |||
@@ -327,6 +327,7 @@ void fiq_playback(void) | |||
327 | */ | 327 | */ |
328 | asm volatile ( | 328 | asm volatile ( |
329 | /* No external calls */ | 329 | /* No external calls */ |
330 | ".syntax unified \n" | ||
330 | "sub lr, lr, #4 \n" /* Prepare return address */ | 331 | "sub lr, lr, #4 \n" /* Prepare return address */ |
331 | "stmfd sp!, { lr } \n" /* stack lr so we can use it */ | 332 | "stmfd sp!, { lr } \n" /* stack lr so we can use it */ |
332 | "ldr r12, =0xcf001040 \n" /* Some magic from iPodLinux ... */ | 333 | "ldr r12, =0xcf001040 \n" /* Some magic from iPodLinux ... */ |
@@ -349,8 +350,8 @@ void fiq_playback(void) | |||
349 | "bhi 0b \n" /* ... yes, continue */ | 350 | "bhi 0b \n" /* ... yes, continue */ |
350 | 351 | ||
351 | "cmp r9, #0 \n" /* either FIFO full or size empty? */ | 352 | "cmp r9, #0 \n" /* either FIFO full or size empty? */ |
352 | "stmneia r11, { r8-r9 } \n" /* save p and size, if not empty */ | 353 | "stmiane r11, { r8-r9 } \n" /* save p and size, if not empty */ |
353 | "ldmnefd sp!, { pc }^ \n" /* RFE if not empty */ | 354 | "ldmfdne sp!, { pc }^ \n" /* RFE if not empty */ |
354 | 355 | ||
355 | /* Making external calls */ | 356 | /* Making external calls */ |
356 | "1: \n" | 357 | "1: \n" |
@@ -363,7 +364,7 @@ void fiq_playback(void) | |||
363 | "mov lr, pc \n" /* long call (not in same section) */ | 364 | "mov lr, pc \n" /* long call (not in same section) */ |
364 | "bx r3 \n" | 365 | "bx r3 \n" |
365 | "cmp r0, #0 \n" /* more data? */ | 366 | "cmp r0, #0 \n" /* more data? */ |
366 | "ldmeqfd sp!, { r0-r3, pc }^ \n" /* no? -> exit */ | 367 | "ldmfdeq sp!, { r0-r3, pc }^ \n" /* no? -> exit */ |
367 | 368 | ||
368 | "ldr r14, [r10, #0x1c] \n" /* read IISFIFO_CFG to check FIFO status */ | 369 | "ldr r14, [r10, #0x1c] \n" /* read IISFIFO_CFG to check FIFO status */ |
369 | "ands r14, r14, #(0xe<<23) \n" /* r14 = (IIS_TX_FREE_COUNT & ~1) << 23 */ | 370 | "ands r14, r14, #(0xe<<23) \n" /* r14 = (IIS_TX_FREE_COUNT & ~1) << 23 */ |
diff --git a/lib/arm_support/support-arm.S b/lib/arm_support/support-arm.S index f99d086b0b..3af0c70975 100644 --- a/lib/arm_support/support-arm.S +++ b/lib/arm_support/support-arm.S | |||
@@ -227,7 +227,7 @@ | |||
227 | /* Test whether divisor is 2^N */ | 227 | /* Test whether divisor is 2^N */ |
228 | cmp \inv, #1<<31 | 228 | cmp \inv, #1<<31 |
229 | /* Load approximate reciprocal */ | 229 | /* Load approximate reciprocal */ |
230 | ldrhib \inv, [\neg, #.L_udiv_est_table-.-64] | 230 | ldrbhi \inv, [\neg, #.L_udiv_est_table-.-64] |
231 | bls 20f | 231 | bls 20f |
232 | subs \bits, \bits, #7 | 232 | subs \bits, \bits, #7 |
233 | rsb \neg, \divisor, #0 | 233 | rsb \neg, \divisor, #0 |
diff --git a/lib/rbcodec/codecs/demac/libdemac/udiv32_arm.S b/lib/rbcodec/codecs/demac/libdemac/udiv32_arm.S index 7b851659bd..1d19160a91 100644 --- a/lib/rbcodec/codecs/demac/libdemac/udiv32_arm.S +++ b/lib/rbcodec/codecs/demac/libdemac/udiv32_arm.S | |||
@@ -225,7 +225,7 @@ udiv32_arm: | |||
225 | mov \inv, \divisor, lsl \bits | 225 | mov \inv, \divisor, lsl \bits |
226 | add \neg, pc, \inv, lsr #25 | 226 | add \neg, pc, \inv, lsr #25 |
227 | cmp \inv, #1<<31 | 227 | cmp \inv, #1<<31 |
228 | ldrhib \inv, [\neg, #.L_udiv_est_table-.-64] | 228 | ldrbhi \inv, [\neg, #.L_udiv_est_table-.-64] |
229 | bls 20f | 229 | bls 20f |
230 | subs \bits, \bits, #7 | 230 | subs \bits, \bits, #7 |
231 | rsb \neg, \divisor, #0 | 231 | rsb \neg, \divisor, #0 |
diff --git a/lib/rbcodec/codecs/demac/libdemac/vector_math16_armv6.h b/lib/rbcodec/codecs/demac/libdemac/vector_math16_armv6.h index 8d27331b62..1da090efbb 100644 --- a/lib/rbcodec/codecs/demac/libdemac/vector_math16_armv6.h +++ b/lib/rbcodec/codecs/demac/libdemac/vector_math16_armv6.h | |||
@@ -45,6 +45,7 @@ static inline int32_t vector_sp_add(int16_t* v1, int16_t* f2, int16_t* s2) | |||
45 | #endif | 45 | #endif |
46 | 46 | ||
47 | asm volatile ( | 47 | asm volatile ( |
48 | ".syntax unified \n" | ||
48 | #if ORDER > 32 | 49 | #if ORDER > 32 |
49 | "mov %[res], #0 \n" | 50 | "mov %[res], #0 \n" |
50 | #endif | 51 | #endif |
@@ -117,7 +118,7 @@ static inline int32_t vector_sp_add(int16_t* v1, int16_t* f2, int16_t* s2) | |||
117 | "smladx %[res], r1, r2, %[res] \n" | 118 | "smladx %[res], r1, r2, %[res] \n" |
118 | #if ORDER > 32 | 119 | #if ORDER > 32 |
119 | "subs %[cnt], %[cnt], #1 \n" | 120 | "subs %[cnt], %[cnt], #1 \n" |
120 | "ldmneia %[f2]!, {r2,r4} \n" | 121 | "ldmiane %[f2]!, {r2,r4} \n" |
121 | "sadd16 r0, r0, r7 \n" | 122 | "sadd16 r0, r0, r7 \n" |
122 | "sadd16 r1, r1, r5 \n" | 123 | "sadd16 r1, r1, r5 \n" |
123 | "strd r0, [%[v1]], #8 \n" | 124 | "strd r0, [%[v1]], #8 \n" |
@@ -172,8 +173,8 @@ static inline int32_t vector_sp_add(int16_t* v1, int16_t* f2, int16_t* s2) | |||
172 | "smlad %[res], r3, r5, %[res] \n" | 173 | "smlad %[res], r3, r5, %[res] \n" |
173 | #if ORDER > 32 | 174 | #if ORDER > 32 |
174 | "subs %[cnt], %[cnt], #1 \n" | 175 | "subs %[cnt], %[cnt], #1 \n" |
175 | "ldrned r4, [%[f2]], #8 \n" | 176 | "ldrdne r4, [%[f2]], #8 \n" |
176 | "ldrned r0, [%[v1], #8] \n" | 177 | "ldrdne r0, [%[v1], #8] \n" |
177 | "sadd16 r2, r2, r6 \n" | 178 | "sadd16 r2, r2, r6 \n" |
178 | "sadd16 r3, r3, r7 \n" | 179 | "sadd16 r3, r3, r7 \n" |
179 | "strd r2, [%[v1]], #8 \n" | 180 | "strd r2, [%[v1]], #8 \n" |
@@ -214,6 +215,7 @@ static inline int32_t vector_sp_sub(int16_t* v1, int16_t* f2, int16_t* s2) | |||
214 | #endif | 215 | #endif |
215 | 216 | ||
216 | asm volatile ( | 217 | asm volatile ( |
218 | ".syntax unified \n" | ||
217 | #if ORDER > 32 | 219 | #if ORDER > 32 |
218 | "mov %[res], #0 \n" | 220 | "mov %[res], #0 \n" |
219 | #endif | 221 | #endif |
@@ -286,7 +288,7 @@ static inline int32_t vector_sp_sub(int16_t* v1, int16_t* f2, int16_t* s2) | |||
286 | "smladx %[res], r1, r2, %[res] \n" | 288 | "smladx %[res], r1, r2, %[res] \n" |
287 | #if ORDER > 32 | 289 | #if ORDER > 32 |
288 | "subs %[cnt], %[cnt], #1 \n" | 290 | "subs %[cnt], %[cnt], #1 \n" |
289 | "ldmneia %[f2]!, {r2,r4} \n" | 291 | "ldmiane %[f2]!, {r2,r4} \n" |
290 | "ssub16 r0, r0, r7 \n" | 292 | "ssub16 r0, r0, r7 \n" |
291 | "ssub16 r1, r1, r5 \n" | 293 | "ssub16 r1, r1, r5 \n" |
292 | "strd r0, [%[v1]], #8 \n" | 294 | "strd r0, [%[v1]], #8 \n" |
@@ -341,8 +343,8 @@ static inline int32_t vector_sp_sub(int16_t* v1, int16_t* f2, int16_t* s2) | |||
341 | "smlad %[res], r3, r5, %[res] \n" | 343 | "smlad %[res], r3, r5, %[res] \n" |
342 | #if ORDER > 32 | 344 | #if ORDER > 32 |
343 | "subs %[cnt], %[cnt], #1 \n" | 345 | "subs %[cnt], %[cnt], #1 \n" |
344 | "ldrned r4, [%[f2]], #8 \n" | 346 | "ldrdne r4, [%[f2]], #8 \n" |
345 | "ldrned r0, [%[v1], #8] \n" | 347 | "ldrdne r0, [%[v1], #8] \n" |
346 | "ssub16 r2, r2, r6 \n" | 348 | "ssub16 r2, r2, r6 \n" |
347 | "ssub16 r3, r3, r7 \n" | 349 | "ssub16 r3, r3, r7 \n" |
348 | "strd r2, [%[v1]], #8 \n" | 350 | "strd r2, [%[v1]], #8 \n" |
@@ -381,6 +383,7 @@ static inline int32_t scalarproduct(int16_t* v1, int16_t* v2) | |||
381 | #endif | 383 | #endif |
382 | 384 | ||
383 | asm volatile ( | 385 | asm volatile ( |
386 | ".syntax unified \n" | ||
384 | #if ORDER > 32 | 387 | #if ORDER > 32 |
385 | "mov %[res], #0 \n" | 388 | "mov %[res], #0 \n" |
386 | #endif | 389 | #endif |
@@ -421,10 +424,10 @@ static inline int32_t scalarproduct(int16_t* v1, int16_t* v2) | |||
421 | "pkhtb r1, r7, r4 \n" | 424 | "pkhtb r1, r7, r4 \n" |
422 | #if ORDER > 32 | 425 | #if ORDER > 32 |
423 | "subs %[cnt], %[cnt], #1 \n" | 426 | "subs %[cnt], %[cnt], #1 \n" |
424 | "ldrned r6, [%[v2]], #8 \n" | 427 | "ldrdne r6, [%[v2]], #8 \n" |
425 | "smladx %[res], r2, r1, %[res] \n" | 428 | "smladx %[res], r2, r1, %[res] \n" |
426 | "pkhtb r2, r4, r5 \n" | 429 | "pkhtb r2, r4, r5 \n" |
427 | "ldrned r0, [%[v1]], #8 \n" | 430 | "ldrdne r0, [%[v1]], #8 \n" |
428 | "smladx %[res], r3, r2, %[res] \n" | 431 | "smladx %[res], r3, r2, %[res] \n" |
429 | "bne 1b \n" | 432 | "bne 1b \n" |
430 | #else | 433 | #else |
@@ -461,9 +464,9 @@ static inline int32_t scalarproduct(int16_t* v1, int16_t* v2) | |||
461 | "ldrd r4, [%[v2]], #8 \n" | 464 | "ldrd r4, [%[v2]], #8 \n" |
462 | "smlad %[res], r1, r6, %[res] \n" | 465 | "smlad %[res], r1, r6, %[res] \n" |
463 | "subs %[cnt], %[cnt], #1 \n" | 466 | "subs %[cnt], %[cnt], #1 \n" |
464 | "ldrned r0, [%[v1]], #8 \n" | 467 | "ldrdne r0, [%[v1]], #8 \n" |
465 | "smlad %[res], r2, r7, %[res] \n" | 468 | "smlad %[res], r2, r7, %[res] \n" |
466 | "ldrned r6, [%[v2]], #8 \n" | 469 | "ldrdne r6, [%[v2]], #8 \n" |
467 | "smlad %[res], r3, r4, %[res] \n" | 470 | "smlad %[res], r3, r4, %[res] \n" |
468 | "bne 1b \n" | 471 | "bne 1b \n" |
469 | #else | 472 | #else |
diff --git a/lib/rbcodec/codecs/libtta/filter_arm.S b/lib/rbcodec/codecs/libtta/filter_arm.S index f3959b83ca..10f1491796 100644 --- a/lib/rbcodec/codecs/libtta/filter_arm.S +++ b/lib/rbcodec/codecs/libtta/filter_arm.S | |||
@@ -165,8 +165,8 @@ hybrid_filter: | |||
165 | sub r10, r11, r10 | 165 | sub r10, r11, r10 |
166 | 166 | ||
167 | @ set to the memory: *pA, *(pA-1), *(pA-2), *(pA-3), *pM, *(pM-1), *(pM-2), *(pM-3) | 167 | @ set to the memory: *pA, *(pA-1), *(pA-2), *(pA-3), *pM, *(pM-1), *(pM-2), *(pM-3) |
168 | stmneda r2, {r10, r11, r12, lr} | 168 | stmdane r2, {r10, r11, r12, lr} |
169 | stmneda r3, {r5, r6, r7, r8} | 169 | stmdane r3, {r5, r6, r7, r8} |
170 | ldmpc cond=ne regs=r4-r12 @ hybrid_filter end (when fs->index != 0) | 170 | ldmpc cond=ne regs=r4-r12 @ hybrid_filter end (when fs->index != 0) |
171 | 171 | ||
172 | .hf_memshl: | 172 | .hf_memshl: |
diff --git a/lib/rbcodec/dsp/dsp_arm.S b/lib/rbcodec/dsp/dsp_arm.S index 16394b8690..864abee4b6 100644 --- a/lib/rbcodec/dsp/dsp_arm.S +++ b/lib/rbcodec/dsp/dsp_arm.S | |||
@@ -323,7 +323,7 @@ resample_hermite: | |||
323 | add r6, r6, r0, lsl #2 @ r6 = &s[pos] | 323 | add r6, r6, r0, lsl #2 @ r6 = &s[pos] |
324 | 324 | ||
325 | cmp r0, #3 @ pos >= 3? history not needed | 325 | cmp r0, #3 @ pos >= 3? history not needed |
326 | ldmgedb r6, { r1-r3 } @ x3..x1 = s[pos-3]..s[pos-1] | 326 | ldmdbge r6, { r1-r3 } @ x3..x1 = s[pos-3]..s[pos-1] |
327 | bge .hrs_loadhist_done @ | 327 | bge .hrs_loadhist_done @ |
328 | add r10, r0, r0, lsl #1 @ branch pc + pos*12 | 328 | add r10, r0, r0, lsl #1 @ branch pc + pos*12 |
329 | add pc, pc, r10, lsl #2 @ | 329 | add pc, pc, r10, lsl #2 @ |
@@ -496,7 +496,7 @@ resample_hermite: | |||
496 | ldmfd sp!, { r10, r12 } @ recover ch, h | 496 | ldmfd sp!, { r10, r12 } @ recover ch, h |
497 | subs r10, r10, #1 @ --ch | 497 | subs r10, r10, #1 @ --ch |
498 | stmia r12!, { r1-r3 } @ h[0..2] = x3..x1 | 498 | stmia r12!, { r1-r3 } @ h[0..2] = x3..x1 |
499 | ldmgtia sp, { r0-r2 } @ load data, src, dst | 499 | ldmiagt sp, { r0-r2 } @ load data, src, dst |
500 | bgt .hrs_channel_loop | 500 | bgt .hrs_channel_loop |
501 | 501 | ||
502 | ldmfd sp!, { r1-r3 } @ pop data, src, dst | 502 | ldmfd sp!, { r1-r3 } @ pop data, src, dst |
@@ -614,7 +614,7 @@ filter_process: | |||
614 | ldr r0, [sp] @ r0 = history[channels-ch-1] | 614 | ldr r0, [sp] @ r0 = history[channels-ch-1] |
615 | subs r3, r3, #1 @ all channels processed? | 615 | subs r3, r3, #1 @ all channels processed? |
616 | stmia r0!, { r9-r12 } @ save back history, history++ | 616 | stmia r0!, { r9-r12 } @ save back history, history++ |
617 | ldmhsib sp, { r1-r2 } @ r1 = buf, r2 = count | 617 | ldmibhs sp, { r1-r2 } @ r1 = buf, r2 = count |
618 | strhs r3, [sp, #12] @ store ch | 618 | strhs r3, [sp, #12] @ store ch |
619 | strhs r0, [sp] @ store history[channels-ch-1] | 619 | strhs r0, [sp] @ store history[channels-ch-1] |
620 | bhs .fp_channelloop | 620 | bhs .fp_channelloop |
diff --git a/lib/rbcodec/dsp/dsp_arm_v6.S b/lib/rbcodec/dsp/dsp_arm_v6.S index a36760f744..aa27ec90f6 100644 --- a/lib/rbcodec/dsp/dsp_arm_v6.S +++ b/lib/rbcodec/dsp/dsp_arm_v6.S | |||
@@ -18,6 +18,7 @@ | |||
18 | * KIND, either express or implied. | 18 | * KIND, either express or implied. |
19 | * | 19 | * |
20 | ****************************************************************************/ | 20 | ****************************************************************************/ |
21 | #include "rbcodecconfig.h" | ||
21 | 22 | ||
22 | /**************************************************************************** | 23 | /**************************************************************************** |
23 | * void sample_output_mono(struct sample_io_data *this, | 24 | * void sample_output_mono(struct sample_io_data *this, |
@@ -56,7 +57,7 @@ sample_output_mono: | |||
56 | stmia r3!, { r12, r14 } @ store So0, So1 | 57 | stmia r3!, { r12, r14 } @ store So0, So1 |
57 | bgt 1b @ | 58 | bgt 1b @ |
58 | @ | 59 | @ |
59 | ldmltfd sp!, { r4, pc } @ if count was even, we're done | 60 | ldmfdlt sp!, { r4, pc } @ if count was even, we're done |
60 | @ | 61 | @ |
61 | 2: @ | 62 | 2: @ |
62 | ldr r12, [r2] @ round, scale, saturate | 63 | ldr r12, [r2] @ round, scale, saturate |
@@ -113,7 +114,7 @@ sample_output_stereo: | |||
113 | stmia r3!, { r6, r7 } @ store So0, So1 | 114 | stmia r3!, { r6, r7 } @ store So0, So1 |
114 | bgt 1b @ | 115 | bgt 1b @ |
115 | @ | 116 | @ |
116 | ldmltfd sp!, { r4-r7, pc } @ if count was even, we're done | 117 | ldmfdlt sp!, { r4-r7, pc } @ if count was even, we're done |
117 | @ | 118 | @ |
118 | 2: @ | 119 | 2: @ |
119 | ldr r6, [r2] @ r6 = Li | 120 | ldr r6, [r2] @ r6 = Li |
diff --git a/lib/unwarminder/safe_read.S b/lib/unwarminder/safe_read.S index 1969c7e0c6..ce4913870c 100644 --- a/lib/unwarminder/safe_read.S +++ b/lib/unwarminder/safe_read.S | |||
@@ -45,7 +45,7 @@ safe_read8_faulty_addr: | |||
45 | @ if(value != NULL) | 45 | @ if(value != NULL) |
46 | cmp r1, #0 | 46 | cmp r1, #0 |
47 | @ *value = r0 | 47 | @ *value = r0 |
48 | strneb r0, [r1] | 48 | strbne r0, [r1] |
49 | @ return true; | 49 | @ return true; |
50 | mov r0, #1 | 50 | mov r0, #1 |
51 | bx lr | 51 | bx lr |
@@ -72,7 +72,7 @@ safe_read16_faulty_addr: | |||
72 | @ if(value != NULL) | 72 | @ if(value != NULL) |
73 | cmp r1, #0 | 73 | cmp r1, #0 |
74 | @ *value = r0 | 74 | @ *value = r0 |
75 | strneh r0, [r1] | 75 | strhne r0, [r1] |
76 | @ return true; | 76 | @ return true; |
77 | mov r0, #1 | 77 | mov r0, #1 |
78 | bx lr | 78 | bx lr |